Does document data extraction work with scanned images, PDFs, or only live camera capture?

Most modern document data extraction systems support multiple input methods, though live camera capture through a mobile SDK is typically the primary and best-supported path, since it allows for real-time capture guidance (glare detection, framing feedback) that static image uploads simply can't offer.

Uploaded images, whether from a phone's photo gallery, a desktop file upload, or a scanned document, are commonly supported as a fallback or alternative input method, useful in situations where live camera capture isn't practical, such as a user completing a process on a desktop computer without a good camera, or a business processing a batch of previously captured document images. The tradeoff is that uploaded images have already been captured under whatever conditions existed at the time, without the benefit of real-time feedback that might have caught a blur or glare issue before submission, which can result in somewhat lower success rates compared to guided live capture.

PDF support varies more by vendor and use case. Some systems accept PDF uploads directly, extracting the embedded image content before running OCR and other extraction processes on it, useful for businesses that receive documents as PDF scans rather than photos, common in some B2B or paperwork-heavy workflows.

Others expect image formats (JPEG, PNG) and would require converting a PDF to an image first, either on the vendor's side or the business's own system, before extraction can proceed.

The practical choice between these input methods usually comes down to the specific use case: remote consumer onboarding through a mobile app strongly favors live camera capture for its guidance and typically higher success rate, while back-office document processing or handling previously collected documents may rely more heavily on upload-based extraction.

ScanDoc supports both live camera capture through its mobile SDKs, with real-time guidance to improve capture quality, and image upload as an alternative input method for scenarios where live capture isn't practical, giving businesses flexibility depending on how and where their documents are actually being submitted.